The goals for this fellowship are: (1) to relate the candidate's existing research training in word learning and sound-category learning to a new linguistic domain, phonological-rule generalization; (2) to evaluate whether similar mechanisms support word learning and phonological-rule generalization, two linguistic domains that have been argued to be qualitatively distinct; in particular, to evaluate whether variability on a dimension irrelevant to the learning task facilitates learning analogously in the two domains. In the word- learning task, variability is introduced in the talkers' voices when infants are habituated to the word-object pairs. In the phonological-rule learning task, infants are familiarized to a language in which 3-syllable words obey a consonant-voicing rule (e.g., the 1st 2 syllables must have the same voicing), and variability is introduced in the vowels. To investigate the mechanism by which variability facilitates learning in the two domains, the distributional structure of the variability in each domain is also manipulated; (3) to develop new research skills for the candidate by exposing her to new language-acquisition topics, distributional learning and phonological-rule learning, and new methodologies, artificial-language methods and the head-turn preference procedure; and (4) to supplement the candidate's research training in these new topics and methods with relevant institutional training at the University of Arizona, both through pertinent courses in speech perception, statistics, and ethics, and active participation in lab meetings, seminars, and colloquia across Psychology, Cognitive Science, and Speech, Language and Hearing Sciences. These goals for the fellowship training, by broadening the candidate's knowledge of language acquisition, expanding her expertise to new methods and theoretical questions, and furthering her publication record, will support her long-term career goals of establishing her own research laboratory and securing a tenure-track professorship at a research university. The proposed research has implications for speech therapy. Current speech therapy methods often fail to fully yield of out-of-clinic generalization, and lack of variability in the speech input provided during therapy may be one limiting factor. Identifying what types of noncriterial variability enhance attention to phonological contrasts and phonological regularities could greatly improve the efficacy of speech therapy. In particular, variability on an irrelevant but salient dimension could potentially be used to draw children's attention to the relevant dimension, improving learning.